Modern recommender systems need to deal with multiple objectives like balancing user engagement with recommending diverse and fresh content. An appealing way to optimally trade these off is by imposing constraints on the ranking according to which items are presented to a user. This results in a constrained ranking optimization problem that can be solved as a linear program (LP). However, off-the-shelf LP solvers are unable to meet the severe latency constraints in systems that serve live traffic. To address this challenge, we exploit the structure of the dual optimization problem to develop a fast solver. We analyze theoretical properties of our solver and show experimentally that it is able to solve constrained ranking problems on synthetic and real-world recommendation datasets an order of magnitude faster than off-the-shelf solvers, thereby enabling their deployment under severe latency constraints.

Millimeter wave (mmWave) communications in the 60 GHz band can provide multi-gigabit rates for emerging bandwidth-intensive applications, and has thus gained considerable interest recently. In this paper, we investigate the problem of efficient scheduling in mmWave wireless personal area networks (WPAN). We develop a frame-based scheduling directional MAC protocol, termed FDMAC, to achieve the goal of leveraging collision-free concurrent transmissions to fully exploit spatial reuse in mmWave WPANs. The high efficiency of FDMAC is achieved by amortizing the scheduling overhead over multiple concurrent, back-to-back transmissions in a row. The core of FDMAC is a graph coloring-based scheduling algorithm, termed greedy coloring (GC) algorithm, that can compute near-optimal schedules with respect to the total transmission time with low complexity. The proposed FDMAC is analyzed and evaluated under various traffic models and patterns. Its superior performance is validated with extensive simulations.

In this paper, we investigate the problem of channel bounding in wireless local area networks (WLANs). We propose a MAC protection mechanism to combat the hidden node problem on non-primary channels of a basic service set (BSS). Simulation results show that the proposed MAC protection mechanism is effective in combating hidden nodes and improving throughput of 80MHz stations. The proposed scheme has been adopted into the IEEE 802.11ac specification as a mandatory feature. Furthermore, we extend the medium access mechanisms defined in 802.11n for a wider bandwidth, evaluate fairness of the 802.11n PIFS medium access mechanisms, and compare the performance of a dynamic bandwidth scheme with that of a static bandwidth implementation.

In this paper, we investigate the problem of multi-user spatial division multiple access (MU SDMA) operation in mmWave wireless networks, within which directional antennas are used to combat the high path loss incurred in the 60GHz band. We study the feasibility of MU SDMA in mmWave networks and propose two MAC protocols to support CSMA/CA based uplink and downlink MU SDMA transmissions. The proposed protocols adopt virtual carrier sensing and allows multiple users to communicate with an access point (AP) simultaneously. Performance analysis and simulation results both show that the proposed protocols can achieve considerable performance improvements over a system that supports only single user (SU) operation.

In recent years, the millimeter wave (mmWave) technology has gained considerable interest due to the huge unlicensed bandwidth (i.e., up to 7GHz) available in the 60GHz band in most part of world. In this paper, we investigate the problem of medium access control (MAC) in mmWave wireless networks, within which directional antennas are used to combat the high path loss incurred in the 60GHz band. We extend a directional CSMA/CA protocol presented in our prior work by exploiting spatial reuse. The proposed protocol adopts virtual carrier sensing and allows non-interfering links to communicate simultaneously. We present a performance analysis as well as simulations to evaluate the proposed protocol. Our results show that the directional MAC with spatial reuse can achieve considerable performance improvements over the 802.11 MAC and the protocol proposed in our prior work. It introduces low protocol overhead and has robust performance even when the network is heavily congested.

Multiple-input multiple-output (MIMO) is one form of the smart antenna technology that uses multiple antennas at both the transmitter and receiver to improve communication performance. In this paper, we investigate the problem of medium access control in wireless local area networks (WLANs) with downlink multi-user MIMO (DL MU MIMO) capability. We propose a CSMA/CA MAC protocol with three response mechanisms for DL MU MIMO and compare the performance of DL MU MIMO with the beam-forming (BF) based approach. A novel per-station weighted queuing mechanism is proposed to mitigate the hidden node problem in the network. Performance analysis and simulation study both show that the proposed DL MU MIMO mechanism incurs low overhead and provides significant throughput performance gain over BF based approach in high SNR scenarios.

In this paper, we investigate the training problem of wireless local area networks (WLANs) with downlink multi-user multiple input multiple output (DL MU MIMO) capability. We extend the 802.11 MAC protocol and propose a few training protocols at the MAC layer to support DL MU MIMO. We provide a capacity analysis based on measurement results from an 802.11n systems, evaluate the overhead of these training protocols, and compare the performance of DL MU MIMO with that of a beam-forming (BF) based approach. Through simulation studies, we find that at high SNR and with implicit training, the DL MU MIMO mechanism provides significant performance gain over the BF based approach. Furthermore, our capacity analysis and simulation study also show that it is critical to define appropriate training intervals based on antenna configuration, channel mobility, and CSI feedback overhead.

In this paper, we investigate the problem of medium access control in mmWave wireless personal area networks (WPAN), within which directional antennas are used to combat the high path loss incurred in the 60GHz frequency band. The conventional CSMA/CA protocol does not work well with directional antennas due to impaired carrier sensing at the transmitters. We explain why existing directional MAC protocols do not work well at 60GHz and propose a novel directional CSMA/CA protocol designed specifically for 60GHz WPANs. Instead of relying on physical carrier sensing, the proposed protocol adopts virtual carrier sensing and relies on a central coordinator to distribute network allocation vector (NAV) information. Both performance analysis and simulation study show that the proposed mechanism incurs low overhead and has robust performance even when the network is heavily congested.

The combination of multi-radio nodes in conjunction with a suitably structured mesh architecture has the potential to solve some of the key limitations of present day mesh networks. We propose and evaluate two practical and self-stabilizing channel assignment algorithms for multi-channel dual-radio mesh backhauls. The objective is to find a channel assignment that maximizes network capacity through balancing traffic load and/or co-channel interference on available channels. Through simulations, we show that the proposed scheme can effectively improve the capacity of wireless mesh networks.

To improve the capacity of wireless ad hoc networks by exploiting multiple available channels, we propose a combined proactive routing and multi-channel medium access control (MAC) protocol. The multi-channel MAC protocol is compatible with IEEE 802.11 MAC and imposes the minimum system requirement among existing multi-channel MAC protocols. Because a proactive routing protocol allows each node to have complete topology information of the network, channel assignment can be closely coupled with a proactive routing protocol. The proposed channel assignment protocol is shown to require fewer channels and exhibit significantly lower communication, computation, and storage complexity than existing channel assignment schemes. We prove the correctness of the proposed channel assignment protocol. In addition, through a performance study, we show that the proposed protocol, by effectively increasing capacity, substantially increases throughput and decreases delay compared to the IEEE 802.11 MAC protocol.

Although it has been an active research area for a number of years, distributed channel assignment remains a challenging problem and existing protocols tend to be complex and usually not suitable for practical implementation. In this paper, we propose three principles that facilitate the design of efficient distributed channel assignment protocols in wireless ad hoc networks. Protocols that implement these design principles are shown to require fewer channels and exhibit significantly lower communication, computation, and storage complexity, compared with existing approaches. As examples, we present two such protocols built on the ad-hoc on-demand distance vector (AODV) routing protocol. In addition, we prove the correctness of the algorithms and derive an upper bound on the number of channels required to both resolve collisions and mitigate interference. Simulation results show that, in many cases, the performance of the proposed protocols can approach that of centralized near-optimal algorithms while maintaining low control overhead.

Despite being the subject of many years of research, distributed channel assignment remains a challenging problem. In this paper, we use a cross-layer approach and present a family of efficient distributed channel assignment protocols for wireless ad hoc networks. These protocols seek to assign distinct channels to each node within all sets of interfering nodes on the same route, subject to a channel availability constraint. Because the proposed protocols combine channel assignment with routing, they exhibit significantly lower communication, computation, and storage complexity than existing channel assignment schemes. We present an example protocol that utilizes the AODV routing protocol. In addition, we prove the correctness of the algorithms and derive an upper bound on the number of channels required to both resolve collisions and mitigate interference. Simulation results show that the performance of the proposed protocols can approach that of centralized near-optimal algorithms when the number of active routes is small.

This paper describes a one-hop self-organized clustering (SOC) algorithm developed specifically for ultra-wideband (UWB) networks. Our research is motivated by the strict regulation of the transmit power of UWB devices, as imposed by U.S. Federal Communications Commission (FCC). To comply with FCC power regulations and to maximize network capacity, we propose a heuristic clustering algorithm to form clusters that have the minimum total power emission subject to constraints on radio range and multiple access capacity. Simulation results show that the proposed algorithm performs better than a random selection algorithm in terms of power emission under all circumstances. Furthermore, the performance of the SOC algorithm can approach that of the k-means algorithm when an appropriate upper limit on the number of nodes in a cluster is imposed.